Intercellular channels permit the direct exchange of ions and small molecules between adjacent cells. These channels are formed from connexins, a family of at least 13 genes. Mutations in connexin genes have been associated with neurodegenerative diseases, deafness, cataracts and cardiovascular abnormalities. The goals of this proposal are to identify the cell or tissue behaviors requiring intercellular communication and to understand why such diversity among connexins is required. Toward this end, we have created lines of mice missing 3 different connexin genes, Cx37, Cx40 and Cx50. Lines lacking Cx37 and Cx40 were created to study the role of communication in vascular endothelia. Unexpectedly, Cx37 knockouts are sterile because ovarian folliculogenesis fails. Cx37 knockouts also display a vascular abnormality in which propagation of vasoconstriction among the vessel wall does not occur. To understand this phenomenon, we will determine the subcellular location of the different vascular connexins in arterioles and test our model, that only Cx37 provides endothelial/smooth muscle communication, with an analysis of the Cx40 knockouts. Cx50 and Cx46 are expressed predominantly in the fibers of the ocular lens which are highly coupled to one another and to lens epithelial cells. Surprisingly, knockouts in either gene develop opacities although loss of Cx46 causes a senile-type cataract while loss of Cx50 leads to a pulverulent type and a lens growth defect. To understand how loss of either connexin leads to cataract formation, we will map the patterns of lens cell communication in wild-type and knockout animals. To determine if problems arise from changes in levels of communication or intrinsic channel properties we will replace the Cx50 coding region with Cx46 creating a targeted gene 'knock-in'. Previously, we showed that loss of communication in the developing Xenopus embryos critically affects gastrulation. However, a variety of correlative evidence suggests that communication could influence much earlier patterning events. We will test this hypothesis by perturbing communication in the early embryos using anti-sense ablation and host- transfer to generate embryos lacking maternally inherited Cx38.